Stitch Fix And The Perils of Subscription Personalization

Based on the business press, you might believe that subscription services are booming. And, for many subscription services, there is an upward trend. Netflix is doing well. Amazon Prime is doing well.

But, according to The Wall Street Journal, subscription clothing brand-business, Stitch Fix is having difficulties. Modern Retail points out that Stitch Fix’ 2023 revenue declined
21 percent Y-O-Y. And, Stitch Fix’s active clients declined 13 percent.

There are multiple problems at Stitch Fix. Many of these problems required cost-cutting including the elimination of full-time stylists. Those are the people who “curate” a customer’s box of items. Some of the problems are due to changed behaviors after Covid. Some of the problems are due to the fashion labels Stitch Fix offers. Some of the problems relate to direct-to-consumer marketing by many of the labels that Stitch Fix’s offers.

But, there is another problem area that goes unmentioned: the perils of personalization. Has technology changed the definition of personalization affecting our expectations?

At its core, Stitch Fix committed to providing customers with highly personalized clothing selected by a personal stylist. Everyone can have their own personal shopper regardless of financial status.

Thinking about personalization at Stitch Fix raises interesting questions: Is it easier to deliver personalization in exercise classes (Peloton), entertainment (Netflix), access to special services unavailable otherwise or without the premium cost (Amazon) and razors (Harry) than clothing? Are the problems at Stitch Fix based on the mandate to satisfy each individual customer’s wants and needs? Can true personalization be delivered by algorithm? Or does true personalization require emotional intuitiveness?

Personalization is different from customization. Customization is transactional. Customization is all about the features and functions. Personalization is experiential. Experiential relates to an occurrence or event making an impression and generating feelings. Technology continues to fuel the delivery of ever more data-based personalized experiences. But, these data-based “experiences” are not based on feelings. These data do not address “How do you feel when you use the service?” Nor do these data address “How do our offerings make you feel?” These data are based on past behavior.

The rise of personalization is having an outsized impact on what we expect from brand-businesses. We expect individualized – personalized – experiences, designed only with “me” specifically in mind. The rise of personalization has changed our understanding of just what personalization should be for each one us as individuals. Personalization is now a necessary part of our buying behaviors.

To illustrate the pervasiveness of personalization, a recent request on Nexis for articles and studies on personalization for the past month turned up 4617 hits. The Wall Street Journal added a special section on AI with an article on how AI can find you the perfect movies, TV shows and books. All the recommendations to the AI are based on past behaviors.

Furthermore, the art of personalization has now morphed into hyper-personalization. Think of this as “personalization on steroids” as one online article reports. Hyper-personalization works to “shape” the customer’s journey with the brand-business.

Hyper-personalization reflects the enormous amount of data available to brand-businesses. But, the data still only reflect behavior, preferences and historic information.

Marriott says that its search engine is undergoing a rebuild. The Marriott rebuild goal is to be hyper-personal providing guests with highly “relevant, seamless and tailored experiences” while generating inspiration, according to Deloitte, the global services group. However, the rebuild consists of more “behavioral data, AI and new search paradigms.” The goal is to create “more rewarding digital experiences.” Yet, inspiration requires emotions and the enhanced search engine will still deal with behaviors.

Marriott is correct in augmenting its search engine. People perceive brand-businesses delivering personalized experiences to be valuable brand-businesses. This is because these personalized brand-business experiences reinforce positive feelings of self, status, respect and uniqueness. Personalization, when it works, can increase brand-business loyalty.

Many subscription other services are built on personalization. Reams of data based on past selections along with algorithms, help subscription services like Spotify, address changing customer needs, wants and problems. Having said this, Spotify is still a “non-profit” brand-business.

Here is another interesting element of personalization that may have a negative effect on Stitch Fix. A recent online analysis indicates that there is a big difference between personal subscriptions for physical offerings and personal subscriptions for digital offerings. Apparently, it is less challenging to recommend and offer movies to customers than it is to recommend and offer clothing. The article points out that physical offerings require more internal logistics and effort. And, depending on the deliverable, that precise physical product must, in most cases, be personalized to one, single person multiple times a month.

Clothing is a very important way in which we can express ourselves and, perhaps, receive external recognition fueling internal recognition. Unlike a movie, TV series, special promotions or dog food, clothing must deliver an individual’s individualized self-perception time and again.

Each time Stitch Fix puts a box of clothing together for a customer, that curated box must satisfy that specific customer’s wants and needs. The curated box must only contain items that the specific customer will desire. Multiply this curation millions of times and the perils of personalization come into focus. At some point, in order to deliver relevantly differentiated experience, the specific customer’s feelings about self and external perceptions come into play.

Stitch Fix’s clothing selections must be perfect for the individual and the individual’s occasions. All the algorithms in the world are still working on matching an item of clothing with an individual’s psyche. Can algorithms based on past purchase behaviors and profiles address these questions? There is no reference to feelings and emotional and social rewards in the 4617 articles on personalization.

One physical personalized deliverable that is doing well is Farmer’s Dog. In this case, the challenge is not really the intended user but the owner of the intended user. Dogs tend to be less fussy when it comes to eating. A friend’s dog once ate an entire bag of Greenies, the dog breath freshener, prompting a stomach pump. Another friend’s dog ate a leather moccasin. Dogs will stand by your feet while you are making a salad, begging for fresh vegetables; waiting for you to drop a cherry tomato or cucumber slice. Fresh food is delicious. Farmer’s Dog delivers fresh food for your dog. The Farmer’s Dog promises to sell freshly prepared, human-grade food. The food is so high-grade and “human-like” that pet owners could eat this food. (As a note, Mars used to make people working on its pet food brand-businesses Pedigree and Whiskas eat the food.)

Yes, you fill out a questionnaire that alerts Farmer’s Dog to the idiosyncrasies of your dog including age, health, etc. A box of your dog’s food arrives with the dog’s name on each package. And, human owners are thrilled when their dog finishes the bowl of fresh food. It may not be that each package of food is individually prepared for every puppy or every overweight dog. But, the impression is that personalization has occurred. Owners love their dog’s reactions.

One could say that Farmer’s Dog benefits from the fact that about 54% of US households have at least one dog. This translates to about 63 million people with dogs. Not all of these dogs eat fresh food even though Farmer’s Dog hopes to alter the fresh food dynamic.

On the other hand, all people wear clothing. It is more difficult to satisfy individuals on an individual basis than puppies.

When someone opens a box from Stitch Fix including a dress, leisure wear or ensemble, there is an expectation of a “Wow” experience of satisfaction and delight that the brand-business knows me. Knowing “me” is complex. And, if I open a box with a dress, leisure wear or ensemble of items, and these items do not match my personal expectations and self-perceptions, I think that the brand-business has not figured out who I am and what I will like. That disappointment can affect future usage and, potentially, brand-business loyalty. Algorithms will need to understand and predict how you feel, not just how, when, where and what you will buy. Stitch Fix can address how a customer feels and what emotions these feelings trigger, the brand-businesses will continue to struggle. Behavior data can only take you so far.

As one analyst told The Wall Street Journal that companies like Stitch Fix, “… need to drive more value to the consumer.” Right now, the new CEO at Stitch Fix is shaking up the organization to focus on ‘’transformation” of the brand-business. why and knowing emotions are what make personalization so marvelous. Right now, Stitch Fix operates on personalization while “expectations of personalization” are far ahead of where Stitch Fix operates. The next generation of AI should tell us how the customer feels and why customers are behaving in specific ways. This will make personalization truly personal and very valuable. Until then, brand-businesses such as Stitch Fix may continue to struggle.